Heterologous Expression of Arabidopsis rty Enhances Drought Tolerance in Strawberry (Fragaria × ananassa Duch.)


 Background Strawberry ( Fragaria × ananassa Duch.) is an important fruit crop worldwide. It was particularly sensitive to drought stress because of their fibrous and shallow root systems. Mutation of Arabidopsis thaliana ROOTY ( RTY ) results in increased endogenous auxin levels and roots and shoot growth, but the effects of this gene in strawberry remain unclear. Results Here, we heterologously expressed Arabidopsis rty in strawberry plants and examined the effects of rty expression on the hormonal and physiological properties of the plants. Heterologous expression of rty induced IAA accumulation and increased the production of adventitious roots as well as trichomes on the abaxial leaf surface of the transgenic plants. Furthermore, the transgenic strawberry plants had increased ABA accumulation and stomatal closure. The transgenic strawberry plants exhibited enhanced water use efficiency and a reduced water loss rate. Additionally, peroxidase and catalase activities were significantly higher in the transgenic plants than in the untransformed controls, and the transgenic plants were more drought tolerant than the wild-type plants. Our results uncover a transgenic approaches can be used to overcome the inherent trade-off between plant growth and drought tolerance by enhancing water use efficiency and reducing water loss rate under water shortage conditions. Conclusions In this study, the rty gene improves hormone-mediated drought tolerance in transgenic strawberry. We demonstrated that the heterologous expression of rty in strawberry improved drought tolerance by promoting auxin and ABA accumulation. These phytohormones together brought about various physiological changes that improved drought tolerance via increased root production, trichome density, and stomatal closure. This study provides the basis for future genetic modifications of strawberry to improve drought tolerance.

The aminotransferase encoded by rty is also responsive to abscisic acid (ABA), which exhibits increased levels due to crosstalk between the auxin and ABA biosynthesis and metabolism pathways [11]. ABA and IAA functionally interact in roots as regulators of growth, development, and tropisms [10,11].
ABA is a classic stress-associated plant hormone that improves tolerance to abiotic stresses [12]. Drought stress is among the most destructive abiotic stresses in the agricultural industry. Plants often experience seasonal water stress due to variable rainfall [13]. ABA is an important modulator of the drought stress response in plants due to its effects on guard cell development, stomatal aperture closure, and the expression of speci c genes associated with drought tolerance [14][15][16][17]. The physiological response of plants to drought directly affects growth, productivity, and survival under water shortage conditions [18][19][20]. Despite the importance of auxin and ABA in the plant's response to drought stress, little is known of how the rty-encoded aminotransferase in uences ABA accumulation, improves drought tolerance, or how variation in ABA accumulation may contribute to environmental adaptation.
Strawberry (Fragaria × ananassa Duch.), a avorful fruit that is popular worldwide, is an important source of various minerals and vitamins [21,22]. The strawberry genome carries eight sets of chromosomes (2n = 8x = 56), which are derived from four diploid ancestors [23]. Because of the superior fruit quality of this hybrid subspecies, it was rapidly distributed around the wide [24]. Cultivated strawberry has grown to be one of the most important fruit crop plants worldwide, with a total annual production of over 8 million tons [25]. Due to their brous and shallow root systems, strawberry plants are particularly sensitive to drought stress. It is challenging to cultivate strawberries in drought-prone regions.
In this study, we analyzed the effect of rty on adventitious root development in strawberry. Our data revealed a novel function for rty related to ABA and drought responses, indicating that this gene has roles beyond the regulation of plant growth and development. The heterologous expression of rty enhances plant tolerance to drought stress via the ABA-mediated regulation of stomatal closure in strawberry. These results have potential signi cance for the cultivation of strawberries in drought-prone regions.

Results
Constitutive heterologous expression of rty increases the ABA and IAA contents In A. thaliana, rty is a critical regulator of endogenous auxin concentration, with consequences for normal growth and development [6,7,9]. The RTY gene is located at 8.89 cM between marker SM114 (8.79 cM) and SGCSNP71 (8.94 cM) on chromosome 2 (Fig. 1a). To determine the regulatory effect of rty on adventitious root development in strawberry, rty was introduced into strawberry via Agrobacterium tumefaciens (Fig. 1b). Putatively transformed shoots were identi ed by screening the culture (Fig. 1c). Seven transgenic lines were con rmed by PCR and western blot analysis (Fig. 1e, f and Fig. S2, S3). The transgenic strawberry plants exhibited strong growth, including broad leaves (Fig. 1d) and an increased number of roots (Fig. 2b,c) and leaf trichomes (Fig. 5b, d), but showed reduced tiller number (Fig. 1d, 1 g) compared to the WT plants.
The rty gene encodes either a transaminase or a C-S lyase involved in IAA biosynthesis [7,9,11]. rty appears to be critical for regulating IAA concentrations [6,7,11]. Thus, we expected the heterologous expression of rty in strawberry to result in IAA accumulation. Therefore, we assayed the endogenous IAA contents on days 0 and 4 in WT and rty transgenic plants cultured on MS medium with no exogenous growth regulators. The concentrations of IAA in the WT were 46.5 ng g − 1 on day 0 and 72.7 ng g − 1 on day 4, but in transgenic plants were 66.0 and 155.3 ng g − 1 respectively. Additionally, the IAA concentrations in the transgenic plants signi cantly increased between the two analyzed time-points, whereas the IAA concentration in the WT plants only slightly increased (Fig. 1h). Thus, quanti cation of IAA concentrations revealed that heterologous expression of rty signi cantly enhanced the accumulation of IAA in the transgenic plants on days 0 and 4.
To clarify whether rty increased ABA levels due to crosstalk between the auxin and ABA biosynthesis and metabolism pathways [11], we next determined the endogenous ABA concentrations. The concentrations of ABA in the WT were 236.3 ng g − 1 on day 0 and 421.7 ng g − 1 on day 4, but in transgenic plants were 543.8 and 963.4 ng g − 1 , respectively. The ABA levels exhibited a similar trend as IAA, with endogenous ABA concentrations being signi cantly higher in the transgenic plants than in the WT (Fig. 1i). These data indicate that the heterologous expression of rty in the transgenic plants stimulates the accumulation of large amounts of IAA and ABA.

High Endogenous Auxin Contents Lead To Early Root Development
Endogenous IAA regulates auxin-dependent developmental processes in plants, including adventitious root formation [6,7,9]. To assess the effects of the increased IAA concentration in the transgenic strawberry plants on root development, we examined the histology of the roots of WT and transgenic plants on MS medium with no exogenous growth regulators. The roots exhibited a developmental pattern common among woody perennials. The meristematic tissues, including the exodermis, cortex, and stele, remained undifferentiated on the rst day. There were no differences on day 0. Differences in root development between the WT and transgenic plants were apparent on day 4. Speci cally, in the transgenic plants at this stage, the arched nature of the xylem poles was lost, the periderm formed from the outer layers of the pericycle, and the outer cell layers containing the exodermis, cortex, and endodermis began to break down and rupture, but these phenomena were only observed on day 5 in WT.
In the WT plants on day 4, the primary xylem, primary phloem, and endodermis began to differentiate, and the vascular cambium formed and gave rise to secondary xylem and phloem tissues. Whereas the roots of the transgenic plants were intact on day 5, in WT plants, the periderm had formed and the exodermis, cortex, and endodermis had ruptured by this time point (Fig. 2a). On day 9, most transgenic plants had 3-5 roots, whereas the WT plants had 1-3 roots (Fig. 2b, c). These results suggest that the high IAA concentrations of the transgenic plants induced early root development and increased the number of roots.
Heterologous expression of rty confers drought tolerance to strawberry ABA accumulation in plants is expected to induce many drought-resistance mechanisms [17,26,27]. To clarify how the increased accumulation of ABA in the transgenic plants enhanced drought tolerance, we compared the effects of a drought treatment on the WT and transgenic plants. Speci cally, we grew the WT and transgenic plants for 2 months in pots and then induced drought conditions by withholding water for 2 weeks. The plants were then rewatered and their growth was monitored for 1 week. The 14-day drought treatment resulted in curled and severely wilted leaves in the WT plants, with many leaves withering and falling off the plants. By contrast, the leaves of the transgenic plants were less affected by the exposure to drought stress, with only a few leaves were curled, wilted, or withered (Fig. 3a). Additionally, 80% of the transgenic plants and 2% of the WT plants survived the 14-day drought treatment (Fig. 3b). These results suggest that the enhanced drought tolerance of the transgenic plants was likely mediated by an ABA-dependent pathway.
Simultaneously, we measured the soil relative water content during drought stress. Whereas the relative soil water content was 100% in the WT and transgenic plants on 0 d (i.e., the day of saturation with water), it was decreased from 19% on day 6 to 5.3% on day 14 in WT, but decreased from 20% on day 6 to 6.6% on day14 in transgenic plants (Fig. S4). These data showed that the soil of transgenic plants was wetter by reducing transpiration.
To further characterize the drought tolerance of the transgenic plants, 30-day-old WT and transgenic plants on subculture medium were treated with various PEG concentrations (0%, 10%, 20%, and 30%) to simulate drought conditions. After 2 days of treatment with 0% PEG, the transgenic plants displayed no obvious differences from the untransformed control plants. However, after two days of treatment, leaves of the WT plants began to roll and wilt in response to the 10% PEG treatment, with the rolling and wilting increasing in severity as the concentration of PEG increased to 30%. These symptoms were most severe for the 30% PEG treatment. By contrast, the leaves of the transgenic plants only began to roll and wilt in response to the 20% PEG treatment. The transgenic plants displayed less rolling and wilting than the WT, even under 30% PEG conditions (Fig. 3c).
The ability of the WT and transgenic plants treated with PEG to scavenge reactive oxygen species (ROS) was assessed by examining the activities of two key antioxidant enzymes (POD and CAT), which are known ROS scavengers. CAT activity was considerably higher in the transgenic plants than in the WT controls. Additionally, the PEG treatment signi cantly enhanced the POD activity of the transgenic plants, ultimately resulting in signi cantly higher POD activity in the transgenic than WT plants (Fig. 3d, e).
To explore the molecular mechanisms underlying the increased drought tolerance of the transgenic plants heterologously expressing rty, we analyzed the expression of stress-responsive genes during the PEG treatment using an RT-qPCR analysis. Speci cally, we analyzed the expression of the following drought-responsive genes, involved in ABA biosynthesis, catabolism, transport, and signaling: NCED3 (nine cis-epoxycarotenoid dioxygenase 3) [28], ABI1 (ABA insensitive 1) [29], RD29A (responsive to dehydration 29) [30], DREB2A (dehydration responsive element-binding protein 2A) [31], and PP2C (type-2C protein phosphatase) [32]. The expression of RD29A (a stress-responsive marker) and DREB2A (a regulator of many water stress-inducible genes) during PEG treatment was upregulated to a greater extent in the transgenic plants than in the WT (Fig. S5). The stress-responsive genes were likewise more responsive to drought stress in the transgenic plants than in the WT, suggesting that stress signals are somehow ampli ed in rty, triggering a stronger drought response (Fig. S5). These results suggest that heterologous expression of rty in strawberry may decrease ROS accumulation by enhancing antioxidant enzyme activities.
The water loss rate is lower in the transgenic plants than in the wild-type plants To investigate the effects of heterologous expression of rty on the physiological status of strawberry plants, we analyzed the water loss rate, water use e ciency, and electrolyte leakage of the transgenic and WT plants. The water loss rate of leaves of one-month-old plants at same position excised from three transgenic plants was lower than that of leaves from WT plants (Fig. 4a). The water use e ciency of leaves from two-month-old transgenic plants grown in pots in the greenhouse was higher than that of the WT (Fig. 4b). Moreover, the electrolyte leakage from fresh leaves was lower in transgenic plants than in the WT (Fig. 4c), likely due to the transgenic plants having less cell membrane damage. The differences in the water loss rate, electrolyte leakage, and water use e ciency between the WT and transgenic plants may contribute to the improved drought tolerance of the transgenic plants heterologously expressing rty.
Transgenic plants heterologously expressing rty have increased ABA-induced stomatal closure Previous research indicates that ABA is an important inducer of stomatal closure, preventing water loss and thereby contributing to drought tolerance [14,33]. We thus compared the stomatal aperture sizes of transgenic and WT plants following ABA treatment. Scanning electron microscopy revealed that the percentage of closed stomata was almost two-fold higher in the transgenic plants than in the WT plants ( Fig. 5a, c). There were no signi cant differences in the average number of stomata between the WT and transgenic plants. Thus, the observed reduced water loss rate and increased water use e ciency of the transgenic plants compared to the WT plants were likely not due to differences in the number of stomata but to variability in stomatal closure (Fig. 5c).
To assess whether ABA induced stomatal closure differently in the transgenic plants, we compared the effects of exogenous ABA addition with that of mock. The average width of transgenic plant stomatal apertures was signi cantly smaller than that of the WT for untreated (CK), mock-treated, or 20-µM ABAtreated plants. However, the width of stomatal apertures was more severely reduced for plants subjected to the 20-µM ABA treatment ( Fig. 5e and Fig. S6). Thus, the transgenic plants heterologously expressing rty exhibited increased ABA-induced stomatal closure.
Trichomes affect the optical properties of the leaf surface and may protect plants from stress damage and reduced water loss through decreasing the rate of transpiration [34][35][36]. In the current study, we revealed that the density and number of epidermal trichomes on the abaxial side of leaves were higher in the transgenic plants than in the WT plants. The abaxial surface per unit area of transgenic and WT leaves were an average of 50 and 30 epidermal trichomes per unit area, respectively (Fig. 5b, d). The greater abundance of epidermal trichomes on the transgenic leaves may have aided in the increase in drought tolerance through reducing water loss and decreasing the rate of transpiration. These results suggest that the increase in drought tolerance is due to an increased number of epidermal trichomes and increased endogenous ABA concentrations, leading to smaller stomates.

Expression of auxin biosynthetic and signaling genes is upregulated in transgenic plants
To determine the molecular mechanisms underlying the phenotypic differences between the WT and transgenic plants, we compared the IAA contents of the WT and transgenic plants during drought treatment. The concentration of IAA in the WT was 31.8 ng g − 1 on day 0, 36.5 ng g − 1 on day 4, and 36.7 ng g − 1 on day 8, but in transgenic plants were 43.0, 46.2, and 40.1 ng g − 1 , respectively. The IAA content was always higher in the transgenic plants than in the WT plants during drought treatment ( Fig. 6). However, these results indicated that the IAA content in the transgenic and WT plants did not increase during drought treatment.
To clarify the mechanism underlying the IAA content difference between the WT and transgenic plants, we examined the expression levels of IAA biosynthetic and signaling genes, including PIN1, AAO1, ARF7, MIX2, YUC1, YUC3, and GA3ox. RT-qPCR analysis indicated that drought treatment upregulated the expression of the IAA biosynthetic and signaling genes in the transgenic plants.
These results suggest that the heterologous expression of rty in transgenic plants induced the expression of IAA biosynthetic and signaling genes, which in turn increased IAA accumulation. Furthermore, the increased ABA levels during drought treatment ( Fig. 6) likely in uence the observed production of additional roots (Fig. 2b, c) and trichomes (Fig. 5b, d) in the transgenic plants.

Expression of stress-inducible genes and ABA biosynthetic genes is upregulated in the transgenic plants
Under drought conditions, ABA concentrations increase to a speci c threshold by midday, inducing ion e ux and inhibiting sugar uptake by guard cells, after which the stomatal apertures decrease in size for the rest of the day [37]. To elucidate the role of ABA during stress responses, we compared the ABA contents of the two-month-old WT and transgenic plants following a drought treatment. The concentration of AAA in the WT were 334.0 ng g − 1 on day 0, 1017.2 ng g − 1 on day 4, and 2635.3 ng g − 1 on day 8, but in transgenic plants were 939.3, 1083.7, and 3471.3 ng g − 1 , respectively. The ABA concentration was signi cantly higher at 8 days after initiating the drought treatment in both the WT and transgenic plants, but was 1.3-fold higher in the transgenic plants (Fig. 7).
To determine the molecular mechanisms underlying this difference in ABA contents between the WT and transgenic plants, we examined the expression of genes involved in ABA biosynthesis, catabolism, transport, and signaling, as well as drought-responsive genes, including the following: NCED3 [28], ABI1 [29], RD29A [30], DREB2A [31], and PP2C [32]. Water de cit stress promotes ABA biosynthesis via the upregulated expression of NCED3 [38]. RT-qPCR analysis indicated that the NCED3 transcript levels were signi cantly higher at 8 days after starting the drought treatment, with the increase being more pronounced in transgenic plants, implying that this gene was actively expressed. Additionally, the expression levels of the ABA-inducible marker genes (RD29A and DREB2A), ABA biosynthetic genes (ABI1, ABA8, and PYL9), a stomatal closure-responsive gene (PP2C), and MYB44 were higher in the transgenic plants than in the WT plants during the drought treatment ( Fig. 7). High PP2C and MYB44 transcript levels may induce stomatal closure [32].
Thus, the heterologous expression of rty in strawberry plants considerably increased ABA accumulation. The expression of stress-inducible genes and ABA biosynthetic genes may then trigger stomatal closure via an ABA-dependent pathway, which may contribute to the observed drought tolerance of the transgenic plants.

Discussion
The rty gene improves hormone-mediated drought tolerance in plants (Fig. 8). In plants, rty promotes auxin and ABA accumulation [10]. The increased auxin concentration induces the production of excess roots and increases the density of leaf trichomes, which would be expected to increase moisture retention. The accumulation of ABA promoted stomatal closure in the plants, which increased the water use e ciency and decreased the water loss rate [39,40]. Plant photosynthesis, respiration, and transpiration are all affected by the rty gene. These changes contributed to the increased tolerance of the plants to drought stress [41][42][43][44][45]. In this study, we demonstrated that the heterologous expression of rty in strawberry improved drought tolerance by promoting auxin and ABA accumulation. These phytohormones together brought about various physiological changes that improved drought tolerance, such as increased root production, trichome density, and stomatal closure.
Heterologous expression of rty increases the drought tolerance of transgenic strawberry plants Transgenic strawberry plants heterologously expressing rty exhibited a strong growth potential (Fig. 1d), and produced more roots (Fig. 2b, c) and leaf trichomes (Fig. 5b, d), but fewer tillers (Fig. 1g), than the WT plants. Measurement of IAA concentrations revealed that the heterologous expression of rty signi cantly increased the IAA levels in transgenic strawberry plants (Fig. 1h). The high IAA contents of the transgenic strawberry plants resulted in earlier root development and increased numbers of roots. These dominant effects were consistent with the general functions of rty in A. thaliana. Moreover, rty expression appears to be critical for regulating IAA concentrations in A. thaliana. A recessive rty mutation also yields high endogenous IAA concentrations. The most extreme phenotypic effects of rty expression are the proliferation of adventitious and lateral roots and the restriction of shoot development. These phenotypes are most likely caused by increases in auxin concentrations [6][7][8][9][10]. Thus, rty plays a critical role in regulating endogenous auxin concentrations to facilitate normal growth and development.
In the current study, we revealed a hitherto unknown function of rty related to drought tolerance, revealing that heterologous expression of this mutant gene increases ABA concentrations in response to drought stress. Thus, rty helps regulate plant growth and development as well as responses to abiotic stress conditions. Speci cally, the heterologous expression of rty signi cantly increased the ABA content of the transgenic plants (Fig. 1i). The transgenic plants had a lower water loss rate, exhibited less electrolyte leakage, and had higher water use e ciency than the untransformed controls (Fig. 4). The observed accumulation of ABA in the transgenic plants in response to drought treatment suggested that the increased drought tolerance of the transgenic plants was mediated by an ABA-dependent pathway.
Previous reports have indicated that ABA is important for stomatal closure, which limits water loss and enhances drought tolerance [14,46]. Because ABA helps regulate stomatal closure [47], we speculate that rty expression maintains narrower stomatal apertures in plants exposed to drought stress, which reduces water loss through transpiration. We determined that the heterologous expression of rty in strawberry likely increases the sensitivity of the transgenic plants to ABA and improves their tolerance to drought stress.
rty expression promotes drought stress responses via ABA-regulated stomatal closure Plant drought responses involve a complex process regulated by multiple molecular and cellular pathways. The expression levels of some genes are upregulated or downregulated by exposure to abiotic stresses, and the heterologous expression of these genes can increase the tolerance of transgenic plants to drought or salt stress [48][49][50][51]. Consistent with this, our results indicated that the eheterologous expression of rty has critical effects on strawberry drought responses. Speci cally, heterologous expression of rty increased the tolerance of the transgenic strawberry plants to drought stress. Moreover, the transgenic plants produced more leaf trichomes and had a higher percentage of closed stomata and smaller stomatal apertures compared to the WT plants ( Fig. 5 and Fig. S6). When plants are subjected to drought stress, some physiological factors (e.g., electrolyte leakage and POD and CAT activities) may be quickly activated to enable these plants to survive under extreme environmental conditions [52][53][54][55]. Thus, physiological indices related to plant osmotic stress caused by drought may be useful measures for quickly and accurately assessing plant resistance to abiotic stress. Electrolyte leakage, which re ects the degree of cell membrane damage [52], was higher in the leaves of WT plants than in those of the transgenic plants (Fig. 4), suggesting that heterologous expression of rty may strengthen the plant cell membrane integrity in response to drought stress. Furthermore, POD and CAT are vital antioxidant enzymes that protect plants from abiotic stress damage [56,57]. In this study, we established that CAT and POD were more active in the transgenic plants than in the WT plants (Fig. 3d, 3e). This information may be useful for clarifying the mechanism underlying the increased drought tolerance of the transgenic plants.
In A. thaliana, many positive and negative regulators have been identi ed and characterized as key components of ABA biosynthesis and drought signaling [14,58]. Biotic stresses upregulate the expression of several ABA biosynthetic genes [59][60][61][62][63][64]. Water stress-induced ABA accumulation is preceded by signi cant increases in Phaseolus vulgaris CED1 transcript and protein levels in the leaves and roots [38,65,66]. In A. thaliana, of the ve NCED genes involved in ABA biosynthesis, the expression of only AtNCED3 is strongly induced by dehydration, although a minor increase in the expression of the other NCED genes has also been reported [38,66,67]. Additionally, AtNCED3 overexpression in transgenic A. thaliana plants increased the ABA content and desiccation tolerance [38]. Similarly, in the current study, some ABA biosynthetic and abiotic stress-responsive genes, including RD29A, DREB2A, NCED3, ABI1, ABA8, PYL9, and PP2C [59][60][61][62][63][68][69][70], were more highly expressed in the transgenic plants than in the WT plants under drought conditions (Fig. 6, 7). These results imply that the increased tolerance of the transgenic plants to drought stress may be in uenced by the upregulated expression of these genes in response to drought conditions. Further study will focus on the how rty interacts with these ABA biosynthetic pathways and with stress-induced signal transduction during the plant's response to drought stress. In addition to providing a foundation for future studies of drought stress tolerance, the results presented here may be used to develop genetically modi ed strawberry varieties with improved drought tolerance.

Conclusion
In this study, we heterologously expressed Arabidopsis rty in strawberry plants. The transgenic strawberry plants had induced IAA accumulation and increased the production of adventitious roots as well as trichomes on the abaxial leaf surface. Also, the ABA accumulation and stomatal closure had increased in the transgenic strawberry plants. It exhibited enhanced water use e ciency, a reduced water loss rate, and more drought tolerant than the wild-type plants. So, our results uncover a transgenic approaches can be used to overcome the inherent trade-off between plant growth and drought tolerance by enhancing water use e ciency and reducing water loss rate under water shortage conditions. This study provides the basis for future genetic modi cations of strawberry to improve drought tolerance.

Methods
Vector construction for heterologous expression of rty A previous study [9] and BLAST analysis (GenBank accession: AY050987) suggested that rty encodes an aminotransferase or a C-S lyase that catalyzes IAA biosynthesis. To functionally characterize rty in strawberry (Fragaria × ananassa), we created an overexpression construct by PCR-based cloning. A cDNA fragment from an A. thaliana mutant (Stock Number CS8156 from Arabidopsis Biological Resource Center) containing the entire rty coding region was ampli ed as previously described [61]. The PCR products and the pBI121 vector were digested with Sma and Sac endonucleases, and the digested PCR and vector products were then ligated with T4 DNA ligase (Promega Corporation, Madison, USA).

Heterologous expression of rty in strawberry plants
Untransformed strawberry (Fragaria × ananassa Duch. cultivar 'Honeoye') was used as the wild type (WT) control for analysis of drought stress. The pBI121-rty binary vector, which harbors a kanamycinresistance gene, was inserted into Agrobacterium tumefaciens GV3101 using a freeze-thaw procedure [71]. Agrobacterium-mediated transformation of 'Honeoye' strawberry was carried out according to the protocol of Jin and Wang [72]. The leaf disc method was used to transform strawberry. The leaves of sterile seedlings subcultured for 25-30 days were cut into 3-5 mm 2 leaf discs. Agrobacterium tumefaciens LBA4404 carrying the target gene was cultured in liquid LB medium containing 100 mg L − 1 kanamycin at 28℃ for 16 h. The leaf discs were placed in this culture for 5 min, transferred to regeneration MS medium (Murashige and Skoog, 1962) containing 6-BA 3.0 mg L −1 and 2, 4 -D 0.1 mg L −1 , cultured in darkness for 1 d; and then transferred to regeneration medium containing 400 mg L −1 cephalosporin and 5 mg L −1 kanamycin, at 25 ± 1 ℃, a light cycle of 16 h, and a light intensity of 30 µmol m −1 s −1 . After 30 days of culture on subculture medium (MS + 6-BA 0.2 mg L −1 + IBA 0.1 mg L −1 ), the shoots of 3 ~ 5 cm in height had formed and these were rooted on root medium (1/2 MS + IBA 0.2 mg L −1 ). Kanamycin-resistant plants were further con rmed by a PCR-based assay using NPTand rty genespeci c primers (Table 1). Western blot analysis was used to verify that the target protein was produced in the transgenic plants, which were then used for subsequent analyses [73]. Table 1 The sequences of the oligonucleotide primers used in this work for vector construct, screening the positive transform, and RT-qPCR.  The concentration of protein in the supernatant was determined using the Bradford assay. All samples were stored at − 70℃. After SDS-PAGE separation, the resolved proteins were electroblotted to a PVDF membrane. Electroblotted membranes were subjected to western blot analysis using anti-rty serum (Beijing Protein Innovation Co., Ltd, China). To increase the serum speci city, the Atrty protein secondary structure, tertiary structure, hydrophobicity, antigenicity, and speci city were analyzed. Recombinantly expressed rty protein (1-131 aa) was used as the immunogen to generate antibody (Fig. S1). Membranes were then treated with alkaline phosphatase-labelled protein and goat anti-rabbit antibody (Jinqiao, Co., The next method was performed as previously described [81].

Drought Treatment
The transgenic and WT plants were subjected to a drought tolerance test. Pots (16 × 16 cm) with the same volume of nutrient soil (peat soil: eld soil: vermiculite at 1 : 1 : 1) per pot. After saturation with 0.8 L of water, the 30-day-old WT and transgenic plants cultured on root medium were planted in pots (one plant per pot). Two-month-old plants were subjected to drought stress by withholding water for 14 days after saturation with 0.8 L of water. About seventy-eight pots of WT and transgenic plants were analyzed per treatment, respectively. The pot positions were often changed to minimize the effects of environmental variability in the greenhouse. Leaves were sampled every 2 days during the drought stress treatment. Three pots were sampled. For the drought treatment, the relative water content was measured every 2 days using a Soil Temperature/Moisture Meter L99-TWS-1(Shanghai Danding International Trade Co., Ltd). Before the drought treatment, the mixed vermiculite and soil was saturated with 0.8 L water. The soil water content before drought was set to 100%, while the relative soil water content after the 14-day drought treatment was ∼6%. Additionally, plants were rewatered after the 14-day drought treatment and the survival rate was calculated 7 days later.

Simulated Drought Stress With PEG 8000
Polyethylene glycol (PEG)-infused medium was prepared as described by Verslues et al [82]. Because PEG cannot be dissolved in agar before pouring medium, PEG-infused medium was prepared by pouring a liquid medium containing PEG on top of the solidi ed agar. An aliquot of 5 mmol L − 1 MES was added to stabilize the pH of the medium (to pH 5.7) and to avoid having to adjust the pH after adding PEG. An appropriate volume of ½ MS containing 15 g L − 1 agar and 5 mmol L − 1 MES was prepared for the agar medium. While the agar medium was still hot, an equal volume of agar medium was divided into the glass asks, autoclaved, and then frozen. An appropriate volume of ½ MS liquid medium containing 5 mmol L − 1 MES (to pH 5.7) was then prepared. Solid PEG 8000 (Sigma Catalog number P-2139) was weighed out at 10%, 20%, 30% (w/v) into the liquid medium after autoclaving, while it was still hot. The PEG-infused liquid medium was lter sterilized with a 0.45 µm lter. The 0.04 L volume of PEG-liquid medium was pipetted on top of the glass asks containing solidi ed agar. The 0.06 L volume of liquid medium was pipetted on top of the glass asks containing solidi ed agar, as the 0 treatment. The agar was added to the PEG liquid medium at a ratio of 2:3. The glass asks were allowed to equilibrate for 24 h at room temperature. Before use, the PEG liquid medium was poured off, careful not to dislodge the agar which may no longer be tightly adhered to the bottom of the glass asks. Thirty-day-old WT and transgenic plants heterologously expressing rty ) on subculture medium were transferred to glass asks containing 0% (CK), 10%, 20%, and 30% PEG-infused medium at 21 °C in a temperature-controlled growth room with a 16-h light/8-h dark photoperiod. Forty-eight hours later, the treated plants were observed, photographed, sampled, and subjected to RT-qPCR and antioxidant enzyme activity analyses.

Gas Exchange Measurements And Water Loss
Photosynthetic parameters were measured as described by Zhao et al. [27]. Three independent plants were analyzed and the experiment was replicated. Gas exchange measurements were taken with the LI-6400 (LI-6400, Li-Cor, USA). Every three mature and fully expanded leaves from three transgenic plants and wild type with good growth status and proper position were randomly sampled. The measurements were taken on two hourly basis from 8:00 to 16:00 in the clear day of April 24-25, 2017. Water Use E ciency (WUE) was calculated the ratio net photosynthetic rate divided by transpiration rate [83]. To assess the water loss, leaves of one-month-old plants from the same position were weighed with an electronic scale, and then placed in a glass culture dish on a layer of lter paper and weighed every hour.

Measurement Of Antioxidant Enzyme Activities
To measure antioxidant enzyme activities, fresh leaves (0.2 g) were cut into small pieces and ground with a mortar and pestle on ice, in a solution comprising 1 mL 0.05 M phosphate buffer (pH 7.8), 3 g polyvinylpyrrolidone, and 0.1 g quartz sand. The mortar was washed twice with 2 mL phosphate buffer (pH 7.8). The resulting solution was poured into a 10-mL centrifuge tube for a nal volume of 7 mL with phosphate buffer (pH 7.8). The tube was centrifuged at 2,500 × g for 20 min at 4 °C. The supernatant was collected, topped up to 10 mL with phosphate buffer (pH 7.8), and used as the enzyme solution for the subsequent analysis of antioxidant enzyme activities. Speci cally, catalase (CAT) activity was assayed at 240 nm. The sample was assayed again 1 min later. The peroxidase (POD) activity in a 0.5-mL aliquot of the enzyme solution was determined at 470 nm. The sample was assayed again 1 min later. The CAT and POD activities were calculated as described by Zhao et al [27].

Reverse Transcription Quantitative PCR Analysis
The expression levels of speci c genes in the WT and transgenic plants were analyzed using a reverse transcription quantitative PCR (RT-qPCR) assay. Total RNA and cDNA were prepared as previously described [84]. RT-qPCR primer sequences are provided in Table 1. Two independent biological replicates and three technical replicates were performed. The resulting data were analyzed according to the 2 −ΔΔCt method as outlined by Livak and Schmittgen [85]. The expression levels of speci c genes were normalized against the Actin expression level.

Declarations
Ethics approval and consent to participate Not applicable.

Consent for publication
Not applicable.

Availability of data and materials
All data generated or analyzed during this study are included in the published article and its supplementary data les (Figs. 1, 2, 3, 4, 5, 6, 7, and 8 and Additional les 1, 2, 3, 4, 5 and 6).

Competing interests
The authors declare that they have no competing interests. rty protein secondary structure, tertiary structure, hydrophobicity, antigenicity, and speci city were analysis. rty protein (1-131aa) recombinant protein expression was used as immunogen to keep away from the protein binding site.
Additional le 2: Figure  Additional le 4: Figure S4 Relative water content during 0-14 days under drought stress in the WT and transgenic plants Relative water content was measured using the soil temperature/moisture meter every 2d during 0-14 days of drought stress. Three biological replicated were performed. Since the mixed vermiculite and soil was saturated with 0.8L water, the relative soil water content before drought was set as 100%. Data are presented as the mean ± SD (n = 3).
Additional le 5: Figure     Data are presented as the mean ± SD (n = 3) (*P < 0.05, Student's t test).  Accumulation of IAA and expression of auxin-related genes Transcript levels were quanti ed by an RT-qPCR assay, and Actin was used as a control. The 30 d old wild-type (WT) and transgenic plants cultured on MS medium that contained 0.2 mg L−1 IBA were planted into the pots (one plant per pot). Two-monthold plants were subjected to drought stress by withholding water for 14 days. Leaves from 0 to 8 days were sampled every 2 days during the drought stress treatment. About seventy-eight pots WT and transgenic plants were performed treatment respectively. Three pots were sampled. The two round screening were carried out. Data are presented as the mean ± SD (n = 3) (*P < 0.05, Student's t test).

Figure 7
Accumulation of ABA and expression of ABA-related genes Transcript levels were quanti ed by an RT-qPCR assay, and Actin was used as the control. The 30 d old wild-type (WT) and transgenic plants cultured on MS medium that contained 0.2 mg L−1 IBA were planted into the pots (one plant per pot).
Two-month-old plants were subjected to drought stress by withholding water for 14 days. Leaves from 0 to 8 days were sampled every 2 days during the drought stress treatment. About seventy-eight pots WT and transgenic plants were performed treatment respectively. Three pots were sampled. The two round screening were carried out. Data are presented as the mean ± SD (n = 3) (*P < 0.05, Student's t test).